I've setup squid cache only volumes for my installation's Squid server [we use
the same machine for fairly decent traffic file services, so I don't want to
risk major fragmentation issues]... I've setup a small cache area which has
about (after all of the directories are created) 18800 free inodes left, which
is about twice over the estimated value needed reported by Squid's startup
messages.
Anyhow... I have 128,xxx free blocks... I've set the disk cache size to 120MB.
I'm figuring those log files may easy grow to a few mb. Am I permitting enough
"slop" here? What is the recommended slop factor? And how well does Squid
handle fs out of space (or inode) errors? Does it take pre-emptive LRU removal
actions at that point, or does it dump core?
=The Brat=
Received on Fri Jan 30 1998 - 22:10:22 MST
This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 16:38:34 MST